+The @cmd{INSERT} command (@pxref{INSERT}) is a more flexible
+alternative to @cmd{INCLUDE}. An @cmd{INCLUDE} command acts the same as
+@cmd{INSERT} with @subcmd{ERROR=STOP CD=NO SYNTAX=BATCH} specified.
+
+The optional @subcmd{ENCODING} subcommand has the same meaning as with @cmd{INSERT}.
+
+@node INSERT
+@section INSERT
+@vindex INSERT
+
+@display
+ INSERT [FILE=]'@var{file_name}'
+ [CD=@{NO,YES@}]
+ [ERROR=@{CONTINUE,STOP@}]
+ [SYNTAX=@{BATCH,INTERACTIVE@}]
+ [ENCODING=@{LOCALE, '@var{charset_name}'@}].
+@end display
+
+@cmd{INSERT} is similar to @cmd{INCLUDE} (@pxref{INCLUDE})
+but somewhat more flexible.
+It causes the command processor to read a file as if it were embedded in the
+current command file.
+
+If @subcmd{CD=YES} is specified, then before including the file, the
+current directory becomes the directory of the included
+file.
+The default setting is @samp{CD=NO}.
+Note that this directory remains current until it is
+changed explicitly (with the @cmd{CD} command, or a subsequent
+@cmd{INSERT} command with the @samp{CD=YES} option).
+It does not revert to its original setting even after the included
+file is finished processing.
+
+If @subcmd{ERROR=STOP} is specified, errors encountered in the
+inserted file causes processing to immediately cease.
+Otherwise processing continues at the next command.
+The default setting is @subcmd{ERROR=CONTINUE}.
+
+If @subcmd{SYNTAX=INTERACTIVE} is specified then the syntax contained in
+the included file must conform to interactive syntax
+conventions. @xref{Syntax Variants}.
+The default setting is @subcmd{SYNTAX=BATCH}.
+
+@subcmd{ENCODING} optionally specifies the character set used by the included
+file. Its argument, which is not case-sensitive, must be in one of
+the following forms:
+
+@table @asis
+@item @subcmd{LOCALE}
+The encoding used by the system locale, or as overridden by the
+@cmd{SET} command (@pxref{SET}). On GNU/Linux and other Unix-like systems,
+environment variables, @i{e.g.}@: @env{LANG} or @env{LC_ALL}, determine the
+system locale.
+
+@item @var{charset_name}
+One of the character set names listed by @acronym{IANA} at
+@uref{http://www.iana.org/assignments/character-sets}. Some examples
+are @code{ASCII} (United States), @code{ISO-8859-1} (western Europe),
+@code{EUC-JP} (Japan), and @code{windows-1252} (Windows). Not all
+systems support all character sets.
+
+@item @code{Auto,@var{encoding}}
+Automatically detects whether a syntax file is encoded in an Unicode
+encoding such as UTF-8, UTF-16, or UTF-32. If it is not, then @pspp{}
+generally assumes that the file is encoded in @var{encoding} (an @acronym{IANA}
+character set name). However, if @var{encoding} is UTF-8, and the
+syntax file is not valid UTF-8, @pspp{} instead assumes that the file
+is encoded in @code{windows-1252}.
+
+For best results, @var{encoding} should be an @acronym{ASCII}-compatible
+encoding (the most common locale encodings are all @acronym{ASCII}-compatible),
+because encodings that are not @acronym{ASCII} compatible cannot be
+automatically distinguished from UTF-8.
+
+@item @code{Auto}
+@item @code{Auto,Locale}
+Automatic detection, as above, with the default encoding taken from
+the system locale or the setting on @subcmd{SET LOCALE}.
+@end table
+
+When ENCODING is not specified, the default is taken from the
+@option{--syntax-encoding} command option, if it was specified, and
+otherwise it is @code{Auto}.
+
+@node OUTPUT
+@section OUTPUT
+@vindex OUTPUT
+@cindex precision, of output
+@cindex decimal places
+
+@display
+OUTPUT MODIFY
+ /SELECT TABLES
+ /TABLECELLS SELECT = [ @var{class}... ]
+ FORMAT = @var{fmt_spec}.
+@end display
+@note{In the above synopsis the characters @samp{[} and @samp{]} are literals.
+They must appear in the syntax to be interpreted.}
+
+@cmd{OUTPUT} changes the appearance of the tables in which results are
+printed. In particular, it can be used to set the format and precision
+to which results are displayed.
+
+After running this command, the default table appearance parameters
+will have been modified and each new output table generated uses
+the new parameters.
+
+Following @code{/TABLECELLS SELECT =} a list of cell classes must
+appear, enclosed in square brackets. This list determines the classes
+of values should be selected for modification. Each class can be:
+
+@table @asis
+@item RESIDUAL
+Residual values. Default: @t{F40.2}.
+
+@item CORRELATION
+Correlations. Default: @t{F40.3}.
+
+@item PERCENT
+Percentages. Default: @t{PCT40.1}.
+
+@item SIGNIFICANCE
+Significance of tests (p-values). Default: @t{F40.3}.
+
+@item COUNT
+Counts or sums of weights. For a weighted data set, the default is
+the weight variable's print format. For an unweighted data set, the
+default is F40.0.
+@end table
+
+For most other numeric values that appear in tables, @code{SET FORMAT}
+may be used to specify the format (@pxref{SET FORMAT}).
+
+The value of @var{fmt_spec} must be a valid output format (@pxref{Input and Output Formats}).
+Note that not all possible formats are meaningful for all classes.
+